START

Most companies don't realize they have a documentation problem until everyone already depends on it.

Customers use it. New hires use it. Engineers use it.

And when documentation falls out of date, the whole system starts working against itself

People stop trusting what they're reading

Teams lose context

And nobody can fix it, because the problem is everywhere at once

The longer it goes unfixed, the harder it is to untangle

That's the opportunity Manicule saw

They're building an AI-native DevRel company for developer tools - owning documentation, technical content, GEO, and distribution across social channels.

The premise isn't that AI should replace expertise. It's that expertise should scale.

Their AI agents audit and test at scale. Their humans own the architecture, the writing, and the creative direction.

Every review improves the system

Every project creates more context. Every iteration raises the bar

What started as a highly manual business - helping developer companies create better technical content - has become a scalable AI-native operation

Today, Manicule works with some of the fastest-growing developer tool companies - including Supermemory, Greptile, and Reducto - has scaled fast over the last few months, and has more demand than it can take on.

Not because they publish more content. Because they help developer companies create content developers actually trust.

🎙️ Naman Bansal & Shreyans Jain Co-founders of Manicule (YC P26) on Fondo START

00:57 What Manicule is building
03:43 The Supermemory customer that changed everything
04:47 Building startups together in high school
05:55 Early pivots, recruiting agencies, and first revenue
06:18 Applying to YC as teenagers
07:24 Lessons from being the first engineer at Supermemory
08:49 How DevRel became a growth engine
09:54 Why long-form content beats viral marketing
10:53 Growing Manicule 6× and crossing $60K MRR
12:17 Why developer documentation is often broken
13:19 How AI changes go-to-market without replacing quality
14:42 The biggest misconception about GEO
16:00 What founders misunderstand about getting into YC
17:42 The hackathon that led to YC
19:05 Turning a manual agency into an AI-native company
21:33 Building toward $5M ARR

learn more at manicule.dev

A production issue shouldn't require a detective
Yet that's how most observability works today

Most observability tools just dump alerts. Duplicates, no context, and you still fix it yourself

Nicolò Magnante  and his cofounder Arseniy Shishaev - who spent years building part of Datadog's metrics product - think that's backwards

So they built Superlog. Observability that's meant never to be opened.

A wizard scans your repo, installs proper OpenTelemetry, and runs daily to keep up as you ship

When something breaks, it groups the errors into one incident, investigates with full context - logs, traces, recent deploys, past Slack threads - and drops a single mergeable PR in Slack.

Merge it
Ignore it
Or open it in Claude Code and tweak it

Vendor-neutral, so you keep every log, trace, and metric - even if you leave

Not another notification. A fix.

🎙️  Nicolò Magnante, CEO & Co-Founder, Superlog "Observability that fixes your bugs"

02:00 Why existing observability tools create friction
03:00 The YC pivot that led to Superlog
05:22 Why observability must evolve beyond monitoring
05:40 Installing observability with a single prompt
06:14 Why startups are the ideal first customers
10:01 Why the future is self-healing software

Learn more at superlog.sh

AI agents can write code, analyze data, and pass the bar exam

They can't order you a pizza because they don't have a phone number

Think about how much of the real world still runs on phone numbers

SMS verification

Delivery coordination

Account creation

Calling a business...

Without one, an agent is stuck behind glass

Manav Modi & cofounder Meet Modi built AgentPhone to fix that

Give your agent a phone number and it can suddenly act

He told his agent to order DoorDash

It created an account, talked to the delivery driver, made the payment

Pizza showed up at his door. 

Y Combinator is a customer. After An engineer tried Twilio and got buried in compliance, AgentPhone had him live in a day

Every person will soon have hundreds, maybe thousands of agents working for them

The hard part isn't making them smart

It's giving them the tools to touch the real world 

‍

🎙️ Manav Modi, CEO & Co-Founder, AgentPhone on Fondo START pod

‍

00:57 Why AI agents need phone numbers to interact with the real world

01:24 The tools that move agents closer to human-level autonomy

01:50 The emerging infrastructure stack powering AI agents

02:18 Why payments may become a critical layer for autonomous agents

02:42 An AI agent ordering DoorDash end-to-end without human involvement

03:08 What happened when 100+ developers built on AgentPhone

03:26 Processing 4,000+ calls and 5,000+ messages in a single day

03:43 Unexpected real-world use cases discovered during the hackathon

04:24 How phone numbers unlock authentication and action

05:37 The vision of every person having hundreds or thousands of agents

06:16 Building identity and authentication systems for AI agents

07:31 The startup pivot that led to AgentPhone and ultimately YC

‍

Check out agentphone.ai

Flowscope’s agents can map an entire department in two weeks.

Then they automate it.

It’s an AI-native consulting firm where agents learn how a business actually runs - process by process - then automate the most manual work directly inside the company’s existing systems

Not in quarters. In days.

Samuel kept seeing the same pattern every time a consulting firm walked into a large organization:

Great strategy
Beautiful slides
Minimal implementation

The recommendations were smart. The operational change rarely came.

He also saw a second problem: Consulting fees were disconnected from outcomes. Clients paid upfront whether the recommendations worked or not.

The incentives were broken by design

So he and his co-founder Javier built the thing consulting always promised but rarely delivered:

Implementation.

The pitch to clients is simple:

Grow your company with the headcount you already have.

🎙️ Samuel Mirpuri, Co-Founder, flowscope on Fondo START pod

00:57 Why traditional consulting stopped at recommendations instead of implementation
02:33 The disconnect between consulting fees and measurable outcomes
03:18 How Flowscope maps entire enterprise departments in two weeks
04:24 Why AI should augment human judgment instead of replacing people
05:30 From aeronautical engineering classmates to AI-native founders
06:02 Sam Altman’s $2M OpenAI token offer to YC startups
07:00 Why AI tokens may become a new startup currency layer
07:29 Customer, Customers, Customers

‍

Check out www.flowscope.com

99% of session replays are your app working exactly as expected (hopefully)

The other 1% is where the friction hides

The bugs.
The failed onboarding flows.
The conversion leaks.

No team has time to watch them all.

That's the problem Michael Egan and his co-founder kept hitting at their last company. They had analytics. They had session replays. They had ideas to improve conversion.

The small fixes always lost to bigger priorities..

So they pivoted and built CodeCanary: An AI product engineer that connects to GitHub, PostHog, and Slack. It watches your session replays, finds the UX friction, and opens pull requests with the fix

Always on. Always shipping.

Not dashboards
Not alerts

Actual fixes

‍

🎙️ Michael Egan, CEO & Co-Founder, CodeCanary on Fondo START pod

‍

00:57 AI agents connected to your product analytics and codebase

01:20 "You wake up to pull requests the next morning"

03:41 Pivoting from enterprise mapping software

04:42 Why nobody actually watches their session replays

05:02 Surfacing the 1% of user behavior that matters

05:47 Autonomous A/B testing tied directly to revenue

06:37 Why this solves problems engineers never had time for

07:30 "With every new user, your product gets a little better"

08:30 How to know when it's time to pivot

‍

Check out www.codecanary.ai

‍

Most cancer drugs target proteins

FinalDose thinks the real opportunity is one layer deeper: DNA

Jeff Liu and his team are building a programmable drug platform that uses genetic mutations to identify and selectively destroy cancer cells

Cells without those mutations are not targeted

Instead of creating a completely new drug for every cancer type, the platform is designed as a reusable system with different genetic instructions

One chassis.
Different disease targets.

The broader bet: DNA is more deterministic than proteins, which could unlock therapeutic approaches traditional drugs struggle to reach

Search.
Destroy.

"A programmable cell elimination platform."

🎙️ Jeff Liu, Cofounder & CEO of FinalDose on START

00:11 Why proteins may be the wrong layer for cancer treatment
01:14 Using DNA mutations as programmable kill signatures
03:20 Building a reusable therapeutic platform instead of one-off drugs
05:00 Why hard biotech problems require interdisciplinary teams
06:00 Getting into YC after a 4 a.m. interview in Japan
08:00 Sam Altman’s $2M AI token offer — and why FinalDose passed
10:08 How AI-native biotech is accelerating programmable medicine
12:00 Why the future of medicine may look more like engineering
13:05 The regulatory bottleneck for programmable therapeutics

Learn more at finaldose.ai

The next big space race won’t be about going up.

It’ll be about bringing things back down.

After four years building satellites at Astranis, Payton Case realized something: as launch costs collapse, the bottleneck flips.

There’s still no infrastructure for manufacturing products in space and returning them safely to Earth.

So Dispatch is building it.

The thesis is simple: gravity is an invisible constraint on manufacturing.

Remove gravity and semiconductor defects drop. Pharmaceutical crystals become more stable. Biological structures that collapse on Earth can solidify in microgravity.

To prove the concept, the team built a full-scale heat shield, drove into the Mojave Desert, and blasted it with a rocket engine at 14× expected re-entry force.

It survived 6× what they needed.

The long-term vision:
Factories in orbit. Permanent industrial infrastructure in space.

🎙️ Payton Case, Co-Founder & CEO of Dispatch, on Fondo START pod

00:12 What Dispatch is building: reusable re-entry vehicles for orbital manufacturing
01:16 The next big space race... will involve bringing things back down to Earth
02:05 Expanding from re-entry vehicles to industrial space stations
02:33 Mojave Desert heat shield testing
03:48 Gravity is this invisible constraint on manufacturing
04:06 Pharmaceutical crystal growth in microgravity
04:53 Semiconductor manufacturing with lower defect rates
06:39 Understanding the modern space infrastructure stack
08:01 3D printing organs in microgravity
09:37 The coolest thing to be done in space has not been thought of yet

Check out dispatch.space

‍

Ansel Dias is using his robot factory to build more robot factories

AutoFAB is a robotic desktop factory that prints, assembles, quality-checks, and packages hardware autonomously

It’s already being used to manufacture more of its own robotic arms

The insight behind it is simple: Prototyping is easy

Production is the completely different ballgame - and that’s where hardware startups die

Factories. QA. Packaging. Scaling production.

That’s the bottleneck AutoFAB wants to eliminate

The long-term vision: AutoFABs making more AutoFABs

🎙️ Ansel Dias, Founder & CEO, AutoFAB on Fondo START pod 

00:18 AutoFAB is a robotic desktop factory
00:33 It prints things, then assembles them
00:41 It does quality checks and even packs shipping boxes
01:03 Putting hardware into production is extremely difficult
02:01 AutoFAB to make more AutoFABs
02:14 Growth will become geometric
02:37 Right now it's making more of its own robotic arms

Check out autofab.net to learn more

The most valuable thing in AI won't be generating answers.

It'll be knowing which ones are right.

Right now AI writes code, solves problems, produces proofs. But there's no way to guarantee any of it is correct. Pedro Nobre is building that guarantee.

Cajal sits at the intersection of formal verification and AI. They use Lean, a language that lets you formalize a statement and derive a proof that's either correct or incorrect. 

Binary. No ambiguity.

The hard part: the space of possible proofs is combinatorially large. Humans somehow navigate it with strange inductive biases. Machines couldn't keep up.

Then reinforcement learning changed what's possible. AI can now iterate against an infinite source of reward: mathematical correctness itself.

The thesis: create a superintelligent mathematician, and you solve most problems.

They're already working with frontier AI labs. Starting in quantum computing and finance. Software verification and cryptography next.

🎙️ Pedro Nobre, Co-Founder, Cajal on Fondo START Pod

‍

01:37 Formal verification explained - verifying whether software or mathematics is correct

02:24 We need to make sure what AI outputs is correct

03:07 Why mathematical proof search is combinatorially difficult

03:42 How reinforcement learning is changing theorem proving

04:11 Why AI is suddenly solving harder math problems

04:28 We already have access to a superhuman mathematician

04:46 The future of checking whether all mathematics is actually correct

05:42 Quantum information theory and applied verification research

06:36 Smart contracts, specifications, and provably correct systems

07:11 If you create a super intelligent mathematician, then you solve most problems.

‍

Check out caj.al

A bug shows up in your product 

Lucent spots it 

Drops an alert in Slack

Someone tags a coding agent in the thread 

It ships a PR

Fixed before anyone knew it was broken

This is the workflow Lucent users are actually running today

Lucent watches your session replays like a human analyst: friction points, bugs, failed upgrades...

Patterns no human has time to sit through

Connects to PostHog. Quick set up. 

Julius, Mastra, BrowserUse and more already on it

Every session replay you're collecting and not watching is data you're wasting

Lucent can help

🎙️ Alisa Rae, Founder, Lucent on Fondo START pod 

00:44 Building an AI that watches your session replays 

1:10 Leaving Australia's Tall Poppy Syndrome for SF 

2:41 Getting rejected by YC as a solo founder 

3:32 Why she decided to keep building alone 

4:36 Raising a $1.3M pre-seed round 

6:43 Pivoting Lucent into AI training data 

7:29 Returning to the original vision 

9:27 Why this was impossible before modern AI 

11:21 AI agents generating PRs from Lucent alerts 

13:28 get a co-founder for YC?

Check out www.lucenthq.com to learn more